Intelligent oral snore-ceasing equipment system based on multi-parameter monitoring and method thereof

Abstract

The invention discloses an intelligent oral snore-ceasing equipment system based on multi-parameter monitoring, and relates to the technical field of intelligent oral snore ceasing. According to the system, physiological data and structural data of a user are collected through a data collection module, and a physiological and structural data set is constructed; the AI dynamic evaluation model module trains an initial model by using a convolutional neural network, continuously optimizes the model and dynamically predicts the snoring symptom risk of an individual by analyzing physiological and structural data; the physiological monitoring module monitors the physiological status of a user in real time, calculates a physiological stability index PSI, judges whether sleep is abnormal or not by combining with a threshold Q1, and adopts an intervention strategy if the sleep is abnormal; the airway disturbance monitoring module judges abnormal breathing by calculating a breathing disturbance coefficient HXY and combining with a threshold Q2, the sleeping posture monitoring module calculates a sleeping posture airway compression coefficient PSRF and combining with a threshold Q3 to judge whether the sleeping posture is reasonable, and if not, a strategy is provided; according to the system, personalized and precise snore stopping intervention can be realized.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent oral anti - snoring, and specifically to an intelligent oral anti - snoring device system based on multi - parameter monitoring. Background Art

[0002] With the acceleration of the pace of life and the general increase in stress levels, the incidence of snoring in the population has been continuously rising. As a common symptom related to sleep breathing, snoring usually manifests as involuntary breathing sounds during sleep. Its root causes are mostly related to factors such as airway narrowing in the throat area, soft tissue collapse, or airflow obstruction. During the occurrence of snoring, when the airflow passes through the restricted upper airway area, it generates vortex disturbances, thereby causing vibrations of structures such as the tongue body, soft palate, and pharyngeal wall muscles, forming snoring sounds of different frequencies and intensities. Some individuals experience short - term apnea during snoring, and in severe cases, it may develop into obstructive sleep apnea - hypopnea syndrome (OSAHS), which has an adverse impact on the cardiovascular system, nervous system, and even the overall sleep quality.

[0003] Currently, the commonly used oral anti - snoring devices in clinical practice are mainly mechanical mandibular advancement devices, which rely on the passive traction of the mandibular structure during user wearing to achieve physical expansion of the throat airway. However, most of these devices have a single structure and passive functions, lacking the ability to dynamically perceive multi - dimensional physiological and structural parameters during the user's sleep process, unable to judge the individual's snoring risk status in real - time, and even less able to achieve active intervention and feedback optimization. Summary of the Invention

[0004] Aiming at the deficiencies of the prior art, the present invention provides an intelligent oral anti - snoring device system based on multi - parameter monitoring to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention is realized through the following technical solutions: An intelligent oral anti - snoring device system based on multi - parameter monitoring, including a data acquisition module, an AI dynamic evaluation model module, a physiological monitoring module, an airway disturbance monitoring module, and a sleeping posture monitoring module;

[0006] The data acquisition module is used to collect the user's tongue pressure, oral temperature, oral muscle group muscle tension, heart rate and heart rate variability, blood oxygen, respiratory rate, and the number of deep breaths to construct a physiological data set; collect head and mandibular posture parameters, sound source offset, air flow smoothness, and throat compression ratio to construct a structural data set;

[0007] The AI dynamic evaluation model module is used to construct an initial model through a convolutional neural network, and use the physiological data set and the structural data set as inputs for training and testing; use the trained model for snoring state evaluation and intelligent intervention decision - making, extract intermediate - layer feature vectors to identify deep - layer features, continuously optimize the model, and realize the dynamic prediction of the snoring risk of different individuals;

[0008] The physiological monitoring module is used to monitor the user's physiological condition in real time, combine the data in the physiological dataset, calculate and obtain the physiological stability index PSI, and compare and analyze it with the first threshold Q1 to determine whether the physiological state of the user is abnormal during sleep. If it is abnormal, a strategy will be given;

[0009] The airway perturbation monitoring module is used to monitor the user's breathing condition in real time, combine the data in the structure dataset, calculate and obtain the respiratory interference coefficient HXY, and compare and analyze it with the second threshold Q2 to determine whether the breathing state of the user is abnormal during sleep. If it is abnormal, a strategy will be given;

[0010] The sleeping posture monitoring module is used to monitor the user's sleeping posture condition in real time, combine the data in the structure dataset, calculate and obtain the sleeping posture airway compression coefficient PSRF, and compare and analyze it with the third threshold Q3 to determine whether the sleeping posture structure of the user is within a reasonable range. If it is not within a reasonable range, a strategy will be given.

[0011] Preferably, the data acquisition module includes a first acquisition unit and a second acquisition unit;

[0012] The first acquisition unit is used to collect the user's tongue pressure TG through a tongue pressure sensor; collect the user's oral temperature value TJ through an oral temperature sensor; collect the muscle tension value Jti of the user's oral muscles through an electromyography sensor; collect the user's heart rate value HR and heart rate variability value HRv through a denture-type heart rate HRV composite sensor; collect the user's blood oxygen value SPO through a blood oxygen micro-optical sensor; collect the user's respiratory rate RR and the number of deep breaths per minute DRR through a respiratory flow detector; and establish a physiological dataset;

[0013] The second acquisition unit is used to collect the user's head elevation angle α, mandibular elevation angle γ, and head offset angle β through a three-axis IMU inertial sensor; collect the sound source posterior fall offset value DTC through a sound source three-dimensional positioning device; collect the impedance ventilation patency Zio through an airflow impedance detector; collect the throat compression ratio Ymc through an impedance change monitor; and establish a structure dataset.

[0014] Preferably, the AI dynamic evaluation model module is used to construct an initial convolutional neural network model for oral snoring monitoring by using a convolutional neural network, and use the physiological dataset and the structure dataset as inputs to train and test the initial convolutional neural network model; use the trained initial convolutional neural network model as the AI dynamic evaluation model for intelligent intervention and regulation, extract the output of the network intermediate layer during the training process as a feature vector for identifying deep features; and further input the extracted feature information into the AI dynamic evaluation model for continuous optimization training and verification to predict the development trend and risk of snoring in different individuals and different states.

[0015] Preferably, the physiological monitoring module includes a first calculation unit and a first analysis unit;

[0016] The first calculation unit is used to monitor the user's physiological condition in real time, combine the data in the physiological data set, and calculate and obtain the physiological stability index PSI after dimensionless processing. The formula is as follows:

[0017] PSI = w1 * HRv + w2 * (1 - HR) + w3 * SPO + w4 * TJ;

[0018] In the formula, HRv represents the user's heart rate variability value, HR represents the user's heart rate value, SPO represents the user's blood oxygen value, TJ represents the user's oral temperature value, and w1, w2, w3, w4, and w5 represent weight coefficients.

[0019] Preferably, the first analysis unit is used to preset a first threshold Q1 in advance, and compare and analyze the physiological stability index PSI with the first threshold Q1 to obtain the first evaluation result, including:

[0020] When the physiological stability index PSI ≥ the first threshold Q1, it indicates that the user's physiological state during sleep is stable, and there is no risk of triggering snoring due to abnormal oxygen deficiency and heart rate fluctuations, and continuous monitoring is carried out;

[0021] When 60% of the first threshold Q1 ≤ the physiological stability index PSI < the first threshold Ql, it indicates that the user's physiological state during sleep is at the first-level abnormality, and there is a risk of triggering snoring due to abnormal oxygen deficiency and heart rate fluctuations. Trigger the first warning instruction and generate the first strategy: use a tongue position muscle movement device to perform low-frequency electrical stimulation on the back of the tongue at 30Hz - 50Hz, 2 seconds each time, so as to reduce the heart rate value by 10% and increase the heart rate variability value by 15%;

[0022] When the physiological stability index PSI < 60% of the first threshold Q1, it indicates that the user's physiological state during sleep is at the second-level abnormality, and there is a risk of triggering snoring due to abnormal oxygen deficiency and heart rate fluctuations. Trigger the second warning instruction and generate the second strategy: monitor the user's breathing condition.

[0023] Preferably, the airway disturbance monitoring module includes a second calculation unit and a second analysis unit;

[0024] When the second calculation unit receives the second warning instruction, it is used to monitor the user's breathing condition in real time, combine the data in the structure data set, and calculate and obtain the breathing interference coefficient HXY after dimensionless processing. The formula is as follows:

[0025]

[0026] Wherein, DTC represents the posterior drop offset value of the sound source, Zio represents the impedance ventilation patency, Jti represents the muscle tension value, Ymc represents the throat compression ratio, RR represents the real-time breathing rate, and RR opt represents the individual optimal breathing rate, DRR represents the number of deep breaths per minute, and DRR max represents the maximum expected number of deep breaths per minute.

[0027] Preferably, the second analysis unit is used to preset a second threshold Q2 in advance, and compare and analyze the breathing interference coefficient HXY with the second threshold Q2 to obtain the second evaluation result, including:

[0028] When the breathing interference coefficient HXY < the second threshold Q2, it indicates that the user's breathing state during sleep is stable, and there is no risk of snoring triggered by structural throat compression or abnormal ventilation function, and continuous monitoring is carried out;

[0029] When the second threshold Q2 * 130% > the breathing interference coefficient HXY ≥ the second threshold Q2, it indicates that the user's breathing state during sleep is at the first level of abnormality, and there is a risk of snoring triggered by structural throat compression or abnormal ventilation function. Trigger the third warning instruction and generate the third strategy: start the micro motor, gently push the tongue body forward by 2 mm; apply a slight vibration stimulation through the flexible diaphragm to activate the sublingual muscle; emit a slow guiding breathing rhythm audio through bone conduction;

[0030] When the breathing interference coefficient HXY ≥ the second threshold Q2 * 130%, it indicates that the user's breathing state during sleep is at the second level of abnormality, and there is a risk of snoring triggered by structural throat compression or abnormal ventilation function. Trigger the fourth warning instruction and generate the fourth strategy: monitor the user's sleeping posture condition.

[0031] Preferably, the sleeping posture monitoring module includes a third calculation unit and a third analysis unit;

[0032] When receiving the fourth warning instruction, the third calculation unit is used to monitor the user's sleeping posture condition in real time, combine the data of the structure data set, and calculate and obtain the sleeping posture airway compression coefficient PSRF after dimensionless processing. The formula is as follows:

[0033]

[0034] Wherein, TG represents the tongue pressure of the user, α represents the head elevation angle, γ represents the mandibular elevation angle, DTC represents the posterior drop offset value of the sound source, Zio represents the impedance ventilation patency, and β avg represents the average head offset angle, and β ref represents the normal posture reference angle.

[0035] Preferably, the third analysis unit is configured to preset a third threshold Q3 in advance, and compare and analyze the sleeping posture airway compression coefficient PSRF with the third threshold Q3 to obtain a third evaluation result, including:

[0036] When the sleeping posture airway compression coefficient PSRF ≤ the third threshold Q3, it indicates that the user's sleeping posture structure is within a reasonable range and there is no risk of airway compression, and continuous monitoring is carried out;

[0037] When the sleeping posture airway compression coefficient PSRF > the third threshold Q3, it indicates that the user's sleeping posture structure is outside the reasonable range and there is a risk of airway compression. Trigger the fifth warning instruction and generate the fifth strategy: use a micro-pneumatic adjustment structure to adjust the mandibular posture by 2-5 degrees, and perform non-awakening intervention with an ultra-low sound pressure <30dB audio and a local bone conduction vibration frequency of 150-250Hz. The initial vibration duration is 0.8 seconds. If the user does not adjust the sleeping posture, it is extended to 2 seconds, and the maximum continuous stimulation does not exceed 5 times.

[0038] Preferably, the intelligent oral snoring stopper method based on multi-parameter monitoring includes the following steps:

[0039] Step 1: Collect the user's tongue pressure, oral temperature, oral muscle tone, heart rate and heart rate variability, blood oxygen, respiratory rate and deep breathing times to construct a physiological data set; collect head and mandibular posture parameters, sound source offset, air flow smoothness and throat compression ratio to construct a structural data set;

[0040] Step 2: Construct an initial model through a convolutional neural network, and use the physiological data set and the structural data set as inputs for training and testing; use the trained model for snoring state evaluation and intelligent intervention decision-making, extract intermediate layer feature vectors to identify deep features, and continuously optimize the model to achieve dynamic prediction of the snoring risk of different individuals;

[0041] Step 3: Real-time monitor the user's physiological condition, combine the data in the physiological data set, calculate and obtain the physiological stability index PSI, and compare and analyze it with the first threshold Q1 to judge whether the user's physiological state is abnormal during sleep. If it is abnormal, a strategy is given;

[0042] Step 4: Real-time monitor the user's breathing condition, combine the data in the structural data set, calculate and obtain the breathing interference coefficient HXY, and compare and analyze it with the second threshold Q2 to judge whether the user's breathing state is abnormal during sleep. If it is abnormal, a strategy is given;

[0043] Step 5: Real-time monitor the user's sleeping posture condition, combine the data in the structural data set, calculate and obtain the sleeping posture airway compression coefficient PSRF, and compare and analyze it with the third threshold Q3 to judge whether the user's sleeping posture structure is within a reasonable range. If it is outside the reasonable range, a strategy is given.

[0044] The present invention provides an intelligent oral snoring stopper system based on multi-parameter monitoring, which has the following beneficial effects:

[0045] (1) The intelligent oral snoring stopper system based on multi-parameter monitoring collects individualized physiological and structural parameters and establishes a dynamic evaluation model in combination with an AI convolutional neural network. It can not only identify deep snoring characteristics, but also continuously train and optimize according to individual differences, realizing accurate prediction and active intervention of snoring risks for different individuals and in different states, and improving the intelligence and personalization levels of snoring prevention.

[0046] (2) The intelligent oral snoring stopper system based on multi-parameter monitoring integrates data from multiple dimensions such as tongue pressure, muscle tension, heart rate variability, blood oxygen, respiratory rate, and head and jaw posture, constructs a physiological data set and a structural data set, and reflects abnormal changes during sleep more comprehensively and precisely than traditional single sensors, improving the real-time recognition ability of problems such as physiological abnormalities, airway disturbances, and improper sleeping postures.

[0047] (3) The intelligent oral snoring stopper system based on multi-parameter monitoring sets multiple evaluation indicators such as physiological stability index, respiratory interference coefficient, and sleeping posture airway compression coefficient, and triggers different levels of warnings through multi-level threshold comparisons, and then automatically invokes diverse intervention methods such as tongue position stimulation, posture adjustment, and bone conduction audio frequency, realizing non-invasive, non-awakening, and hierarchical progressive intervention control, and enhancing user comfort and system response flexibility.

[0048] (4) The intelligent oral snoring stopper system based on multi-parameter monitoring forms a closed-loop operation mechanism of perception, recognition, decision-making, intervention, and verification through continuous monitoring, evaluation, and feedback adjustment, enabling the snoring stopper to not only have a passive support function, but also have an active perception and regulation ability, effectively improving the intelligent response level and overall intervention efficiency of the system, and being applicable to long-term individualized health management scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 It is a block diagram flow chart of the intelligent oral snoring stopper system based on multi-parameter monitoring of the present invention;

[0050] Figure 2 It is a schematic diagram of the method steps of the intelligent oral snoring stopper based on multi-parameter monitoring of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0051] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0052] Example 1

[0053] Please refer to Figure 1 , the present invention provides an intelligent oral snoring stopper system based on multi-parameter monitoring, including a data acquisition module, an AI dynamic evaluation model module, a physiological monitoring module, an airway perturbation monitoring module, and a sleeping posture monitoring module;

[0054] The data acquisition module is used to collect the tongue pressure, oral temperature, muscle tension of oral muscles, heart rate and heart rate variability, blood oxygen, respiratory rate, and the number of deep breaths of the user to construct a physiological data set; collect the head and mandibular posture parameters, sound source offset, airway patency, and throat compression ratio to construct a structural data set;

[0055] The AI dynamic evaluation model module is used to construct an initial model through a convolutional neural network, and use the physiological data set and the structural data set as inputs for training and testing; use the trained model for snoring state evaluation and intelligent intervention decision-making, extract intermediate layer feature vectors to identify deep features, and continuously optimize the model to achieve dynamic prediction of the snoring risk of different individuals;

[0056] The physiological monitoring module is used to monitor the physiological condition of the user in real time, combine the data in the physiological data set, calculate and obtain the physiological stability index PSI, and compare and analyze it with the first threshold Q1 to judge whether the physiological state of the user during sleep is abnormal. If it is abnormal, a strategy is given;

[0057] The airway perturbation monitoring module is used to monitor the breathing condition of the user in real time, combine the data in the structural data set, calculate and obtain the respiratory interference coefficient HXY, and compare and analyze it with the second threshold Q2 to judge whether the breathing state of the user during sleep is abnormal. If it is abnormal, a strategy is given;

[0058] The sleeping posture monitoring module is used to monitor the sleeping posture condition of the user in real time, combine the data in the structural data set, calculate and obtain the sleeping posture airway compression coefficient PSRF, and compare and analyze it with the third threshold Q3 to judge whether the sleeping posture structure of the user is within a reasonable range. If it is not within a reasonable range, a strategy is given.

[0059] In this embodiment, by integrating multi-source sensing and intelligent analysis modules, multi-dimensional real-time monitoring of key factors such as the user's tongue pressure, physiological parameters, head and jaw posture, and throat structure is realized, and the snoring risk is dynamically evaluated based on a convolutional neural network. Layer-by-layer analysis is performed from the physiological state, breathing condition to the sleeping posture structure, and a hierarchical early warning and corresponding intervention mechanism are constructed, significantly improving the personalization, intelligence, and intervention accuracy of the snoring stopper system, and effectively improving the user's sleep quality.

[0060] Example 2

[0061] This embodiment is an explanatory description based on Embodiment 1. Specifically, the data acquisition module includes a first acquisition unit and a second acquisition unit;

[0062] The first acquisition unit is used to collect the user's tongue pressure TG through a tongue pressure sensor; collect the user's oral temperature value TJ through an oral temperature sensor; collect the muscle tension value Jti of the user's oral muscles through an electromyography sensor; collect the user's heart rate value HR and heart rate variability value HRv through a denture-type heart rate HRV composite sensor; collect the user's blood oxygen value SPO through a blood oxygen micro-optical sensor; collect the user's respiratory rate RR and the number of deep breaths per minute DRR through a respiratory flow detector; and establish a physiological data set;

[0063] The second acquisition unit is used to collect the user's head elevation angle α, mandibular elevation angle γ, and head offset angle β through a three-axis IMU inertial sensor; collect the sound source posterior fall offset value DTC through a sound source three-dimensional positioning device; collect the impedance ventilation patency Zio through an airflow impedance detector; collect the throat compression ratio Ymc through an impedance change monitor; and establish a structural data set.

[0064] In this embodiment, by setting the first acquisition unit and the second acquisition unit, a comprehensive perception of the user's physiological parameters and head and neck structure status is realized, and a physiological data set and a structural data set are respectively constructed, providing a high-precision and multi-dimensional data basis for subsequent AI model analysis and personalized intervention, thereby improving the accuracy of snoring monitoring and the timeliness of intervention response.

[0065] Embodiment 3

[0066] This embodiment is an explanatory description based on Embodiment 2. Specifically, the AI dynamic evaluation model module is used to construct an initial convolutional neural network model for oral snoring monitoring using a convolutional neural network, and use the physiological data set and the structural data set as inputs to train and test the initial convolutional neural network model; use the trained initial convolutional neural network model as an AI dynamic evaluation model for intelligent intervention regulation, extract the output of the network intermediate layer during training as a feature vector for identifying deep features; and further input the extracted feature information into the AI dynamic evaluation model for continuous optimization training and verification to predict the development trend and risk of snoring in different individuals and different states.

[0067] In this embodiment, by constructing an AI dynamic evaluation model module based on a convolutional neural network and combining the physiological data set and the structural data set for training and optimization, the accurate identification of deep features of snoring and the prediction of dynamic trends are realized, effectively improving the system's ability to judge individual snoring risks and the scientificity and adaptability of intelligent intervention.

[0068] Embodiment 4

[0069] This embodiment is an explanatory description carried out in Embodiment 3. Specifically, the physiological monitoring module includes a first calculation unit and a first analysis unit;

[0070] The first calculation unit is used to monitor the user's physiological condition in real time. After dimensionless processing in combination with the data in the physiological data set, the physiological stability index PSI is calculated and obtained. The formula is as follows:

[0071] PSI = w1*HRv + w2*(1 - HR) + w3*SPO + w4*TJ;

[0072] In the formula, HRv represents the user's heart rate variability value, HR represents the user's heart rate value, SPO represents the user's blood oxygen value, TJ represents the user's oral temperature value, and w1, w2, w3, w4, and w5 represent weight coefficients, and their specific values are adjusted and set by the user. 0 ≤ w1 ≤ 1, 0 ≤ w2 ≤ 1, 0 ≤ w3 ≤ 1, 0 ≤ w4 ≤ 1, and w1 + w2 + w3 + w4 = 1.

[0073] In this embodiment, the first calculation unit in the physiological monitoring module performs dimensionless processing on multiple physiological parameters and calculates the physiological stability index PSI, which can realize the quantitative evaluation of the user's physiological state during sleep and provide a scientific basis for subsequent abnormal detection and personalized intervention strategies.

[0074] Embodiment 5

[0075] This embodiment is an explanatory description carried out in Embodiment 4. Specifically, the first analysis unit is used to preset a first threshold Q1 in advance and compare and analyze the physiological stability index PSI with the first threshold Q1 to obtain the first evaluation result, including:

[0076] When the physiological stability index PSI ≥ the first threshold Q1, it indicates that the user's physiological state during sleep is stable, without the risk of triggering snoring due to abnormal oxygen deficiency and heart rate fluctuations, and continuous monitoring is carried out;

[0077] When 60% of the first threshold Q1 ≤ the physiological stability index PSI < the first threshold Q1, it indicates that the user's physiological state during sleep is at the first-level abnormality, with the risk of triggering snoring due to abnormal oxygen deficiency and heart rate fluctuations. The first warning instruction is triggered, and the first strategy is generated: through the tongue position muscle movement device, using 30 Hz - 5 Hz, 2 seconds each time, to perform low-frequency electrical stimulation on the back of the tongue, thereby reducing the heart rate value by 10% and increasing the heart rate variability value by 15%;

[0078] When the physiological stability index PSI < 60% of the first threshold Q1, it indicates that the user's physiological state during sleep is at the second-level abnormality, with the risk of triggering snoring due to abnormal oxygen deficiency and heart rate fluctuations. The second warning instruction is triggered, and the second strategy is generated: monitor the user's breathing condition.

[0079] In this embodiment, the physiological stability index PSI is hierarchically judged against a preset threshold Q1 by the first analysis unit, and corresponding stratified early warning and intervention strategies are triggered, enabling early identification and precise response to snoring risks. Especially in the first-level abnormal state, the low-frequency electrical stimulation of the dorsal tongue can improve heart rate variability and reduce heart rate, effectively intervening in potential physiological abnormalities and enhancing the timeliness and individual adaptability of anti-snoring intervention.

[0080] Embodiment 6

[0081] This embodiment is an explanatory description based on Embodiment 5. Specifically, the airway disturbance monitoring module includes a second calculation unit and a second analysis unit;

[0082] The second calculation unit is used to monitor the user's breathing condition in real time when receiving the second early warning instruction. After dimensionless processing in combination with the data in the structural data set, the breathing interference coefficient HXY is calculated, and the formula is as follows: <##

[0083]

[0084] In the formula, DTC represents the numerical value of the posterior displacement of the sound source, Zio represents the impedance ventilation patency, Jti represents the muscle tension value, Ymc represents the throat compression ratio, RR represents the real-time breathing frequency, RR opt represents the individual optimal breathing frequency, DRR represents the number of deep breaths per minute, DRR max represents the maximum expected number of deep breaths per minute.

[0085] In this embodiment, after receiving the second early warning instruction, the second calculation unit calculates the breathing interference coefficient HXY by integrating the structural data set and physiological parameters, which can quantitatively measure the disturbance degree of the user's respiratory system in multiple dimensions, achieve precise assessment of the abnormal breathing state, provide a scientific basis for subsequent intelligent intervention, and enhance the sensitivity and response efficiency of the overall system in snoring risk control.

[0086] Embodiment 7

[0087] This embodiment is an explanatory description based on Embodiment 6. Specifically, the second analysis unit is used to preset a second threshold Q2 in advance, and compare and analyze the breathing interference coefficient HXY with the second threshold Q2 to obtain the second evaluation result, including:

[0088] When the breathing interference coefficient HXY < the second threshold Q2, it indicates that the user's breathing state during sleep is stable, and there is no risk of snoring triggered by structural throat compression or abnormal ventilation function, and continuous monitoring is carried out;

[0089] When the second threshold Q2 * 130% > the respiratory interference coefficient HXY ≥ the second threshold Q2, it indicates that the user's respiratory state is at the first - level abnormality during sleep, there is a risk of triggering snoring due to structural throat compression or abnormal ventilation function, triggering the third warning instruction, and generating the third strategy: starting the micro - motor to slightly push the tongue forward by 2 mm; applying a subtle vibration stimulation through the flexible diaphragm to activate the sublingual muscles; emitting a slow breathing - rhythm - guiding audio through bone conduction;

[0090] When the respiratory interference coefficient HXY ≥ the second threshold Q2 * 130%, it indicates that the user's respiratory state is at the second - level abnormality during sleep, there is a risk of triggering snoring due to structural throat compression or abnormal ventilation function, triggering the fourth warning instruction, and generating the fourth strategy: monitoring the user's sleeping posture.

[0091] In this embodiment, by using the second analysis unit to perform a hierarchical comparison and analysis of the respiratory interference coefficient HXY with the preset threshold Q2, the intelligent judgment of the degree of respiratory abnormality can be realized, and according to the evaluation level, personalized intervention strategies such as tongue forward - pushing, vibration stimulation, and rhythm - audio guidance can be accurately triggered, effectively alleviating the problems of ventilation disorders and throat compression, thereby reducing the probability of triggering snoring.

[0092] Embodiment 8

[0093] This embodiment is an explanatory description based on Embodiment 7. Specifically, the sleeping - posture monitoring module includes a third calculation unit and a third analysis unit;

[0094] The third calculation unit is used to, when receiving the fourth warning instruction, perform real - time monitoring of the user's sleeping posture, combine the data of the structure data set, and after dimensionless processing, calculate and obtain the sleeping - posture airway compression coefficient PSRF. The formula is as follows:

[0095]

[0096] In the formula, TG represents the tongue pressure of the user, which refers to the vertical pressure of the tongue body on the intelligent oral anti - snoring device, α represents the head elevation angle, γ represents the mandibular elevation angle, DTC represents the sound - source posterior - fall offset value, Zio represents the impedance ventilation patency, β avg represents the average head - offset angle, and β ref represents the normal - posture reference angle.

[0097] In this embodiment, by the third calculation unit performing real - time calculation on the sleeping - posture - related parameters to obtain the sleeping - posture airway compression coefficient PSRF, the influence of the user's sleeping posture on airway patency can be comprehensively quantified, realizing the dynamic fusion evaluation of multi - dimensional structural factors such as tongue pressure, head and mandibular angles, and sound - source offset, thereby providing a scientific basis for subsequent personalized posture intervention and effectively reducing the airway compression and snoring risk caused by bad sleeping postures.

[0098] Embodiment 9

[0099] This embodiment is explained in Embodiment 8. Specifically, the third analysis unit is used to preset a third threshold Q3 in advance, and compare and analyze the sleep posture airway compression coefficient PSRF with the third threshold Q3 to obtain the third evaluation result, including:

[0100] When the sleep posture airway compression coefficient PSRF ≤ the third threshold Q3, it indicates that the user's sleep posture structure is within a reasonable range and there is no risk of airway compression, and continuous monitoring is carried out;

[0101] When the sleep posture airway compression coefficient PSRF > the third threshold Q3, it indicates that the user's sleep posture structure is outside the reasonable range and there is a risk of airway compression. Trigger the fifth warning instruction and generate the fifth strategy: use a micro pneumatic adjustment structure to adjust the mandibular posture by 2 - 5 degrees, and perform non-awakening intervention with an ultra-low sound pressure < 30dB audio and a local bone conduction vibration frequency of 150 - 250Hz. The initial vibration duration is 0.8 seconds. If the user does not adjust the sleep posture, it is extended to 2 seconds, and the maximum continuous stimulation does not exceed 5 times.

[0102] In this embodiment, by comparing and analyzing the sleep posture airway compression coefficient PSRF with the preset threshold Q3 through the third analysis unit, the intelligent determination of the rationality of the user's sleep posture can be realized. When an abnormal sleep posture structure is detected, the fifth strategy is triggered for non-awakening intervention. The mandibular posture is adjusted using a micro pneumatic structure and combined with low sound pressure audio and bone conduction vibration to accurately guide the user to finely adjust the sleep posture and effectively relieve the airway compression problem.

[0103] Embodiment 10

[0104] The intelligent oral snore stopper method based on multi-parameter monitoring, please refer to Figure 2 , including the following steps:

[0105] Step 1: Collect the user's tongue pressure, oral temperature, oral muscle tone, heart rate and heart rate variability, blood oxygen, respiratory rate and deep breathing times to construct a physiological data set; collect the head and mandibular posture parameters, sound source offset, air flow smoothness and throat compression ratio to construct a structural data set;

[0106] Step 2: Construct an initial model through a convolutional neural network, and use the physiological data set and the structural data set as inputs for training and testing; use the trained model for snoring state evaluation and intelligent intervention decision-making, extract the intermediate layer feature vectors to identify the deep features, and continuously optimize the model to realize the dynamic prediction of the snoring risk of different individuals;

[0107] Step 3: Real-time monitor the user's physiological condition, combine the data in the physiological data set, calculate and obtain the physiological stability index PSI, and compare and analyze it with the first threshold Q1 to judge whether the user's physiological state is abnormal during sleep. If it is abnormal, a strategy is given;

[0108] Step 4: Monitor the user's breathing status in real time, calculate and obtain the breathing interference coefficient HXY by combining the data in the structure dataset, and compare and analyze it with the second threshold Q2 to determine whether the breathing status of the user is abnormal during sleep. If it is abnormal, a strategy will be given.

[0109] Step 5: Monitor the user's sleeping posture status in real time, calculate and obtain the posture airway compression coefficient PSRF by combining the data in the structure dataset, and compare and analyze it with the third threshold Q3 to determine whether the user's sleeping posture structure is within a reasonable range. If it is not within a reasonable range, a strategy will be given.

[0110] In this embodiment, by comprehensively collecting the physiological and structural data of the user through the steps and combining with the neural network model, the sleep status of the user can be monitored in real time, and the intelligent intervention strategy can be dynamically adjusted. Through the real-time calculation and analysis of key indicators such as physiological stability, breathing interference, and sleeping posture structure, when an abnormality is detected, the system can automatically trigger targeted intervention measures to effectively avoid the occurrence and development of snoring, thereby improving the user's sleep quality and health level.

[0111] The setting of the threshold value is for the convenience of comparison. Regarding the size of the threshold value, it depends on the amount of sample data and the base quantity set by those skilled in the art for each group of sample data; as long as the proportional relationship between the parameter and the quantized value is not affected.

[0112] The above formulas are all obtained by collecting a large amount of data for software simulation and selecting a formula close to the true value. The coefficients in the formulas are set by those skilled in the art according to the actual situation. The above is only a preferred specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention, according to the technical solution and inventive concept of the present invention, makes equivalent substitutions or changes, and should be covered by the protection scope of the present invention.

Claims

1. An intelligent oral anti-snoring device system based on multi-parameter monitoring, characterized in that It includes a data acquisition module, an AI dynamic evaluation model module, a physiological monitoring module, an airway perturbation monitoring module, and a sleeping posture monitoring module; The data acquisition module is used to collect the user's tongue pressure, oral temperature, muscle tension of oral muscles, heart rate and heart rate variability, blood oxygen, respiratory rate, and number of deep breaths, and construct a physiological data set; collect head and mandibular posture parameters, sound source offset, airway patency, and throat compression ratio, and construct a structural data set; The AI dynamic evaluation model module is used to construct an initial model through a convolutional neural network, and use the physiological data set and the structural data set as inputs for training and testing; use the trained model for snoring state evaluation and intelligent intervention decision-making, extract intermediate layer feature vectors to identify deep features, and continuously optimize the model to achieve dynamic prediction of the snoring risk of different individuals; The physiological monitoring module is used to monitor the user's physiological condition in real time, combine the data in the physiological data set, calculate and obtain the physiological stability index PSI, and compare and analyze it with the first threshold Q1 to judge whether the user's physiological state is abnormal during sleep. If it is abnormal, a strategy will be given; The airway perturbation monitoring module is used to monitor the user's breathing condition in real time, combine the data in the structural data set, calculate and obtain the respiratory interference coefficient HXY, and compare and analyze it with the second threshold Q2 to judge whether the user's breathing state is abnormal during sleep. If it is abnormal, a strategy will be given; The sleeping posture monitoring module is used to monitor the user's sleeping posture condition in real time, combine the data in the structural data set, calculate and obtain the sleeping posture airway compression coefficient PSRF, and compare and analyze it with the third threshold Q3 to judge whether the user's sleeping posture structure is within a reasonable range. If it is not within a reasonable range, a strategy will be given.

2. The intelligent oral snore stopper system based on multi-parameter monitoring according to claim 1, characterized in that The data acquisition module includes a first acquisition unit and a second acquisition unit; The first acquisition unit is used to collect the user's tongue pressure TG through a tongue pressure sensor; collect the user's oral temperature value TJ through an oral temperature sensor; collect the muscle tension value Jti of the user's oral muscles through an electromyography sensor; collect the user's heart rate value HR and heart rate variability value HRv through a denture-type heart rate HRV composite sensor; collect the user's blood oxygen value SPO through a blood oxygen micro-optical sensor; collect the user's respiratory rate RR and the number of deep breaths per minute DRR through a respiratory flow detector; and establish a physiological data set; The second acquisition unit is used to collect the user's head elevation angle α, mandibular elevation angle γ, and head offset angle β through a three-axis IMU inertial sensor; collect the sound source posterior fall offset value DTC through a sound source three-dimensional positioning device; collect the impedance ventilation patency Zio through an airflow impedance detector; collect the throat compression ratio Ymc through an impedance change monitor; and establish a structural data set.

3. The intelligent oral snore stopper system based on multi-parameter monitoring according to claim 2, characterized in that, The AI dynamic evaluation model module is used to use a convolutional neural network to construct an initial convolutional neural network model for oral snoring monitoring, and use the physiological data set and the structural data set as inputs to train and test the initial convolutional neural network model; use the trained initial convolutional neural network model as an AI dynamic evaluation model for intelligent intervention regulation, and extract the output of the middle layer of the network during training as a feature vector for identifying deep features; The extracted feature information is further input into the AI dynamic evaluation model for continuous optimization training and verification to predict the development trend and risk of snoring in different individuals and different states.

4. The intelligent oral snore stopper system based on multi-parameter monitoring according to claim 3, characterized in that, The physiological monitoring module includes a first calculation unit and a first analysis unit; The first calculation unit is used to monitor the user's physiological condition in real time. After dimensionless processing in combination with the data in the physiological dataset, the physiological stability index PSI is calculated and obtained. The formula is as follows: PSI = w1*HRv + w2*(1 - HR) + w3*SPO + w4*TJ; In the formula, HRv represents the user's heart rate variability value, HR represents the user's heart rate value, SPO represents the user's blood oxygen value, TJ represents the user's oral temperature value, and w1, w2, w3, w4, and w5 represent weight coefficients.

5. The intelligent oral snore stopper system based on multi-parameter monitoring according to claim 4, characterized in that, The first analysis unit is used to preset a first threshold Q1 in advance and compare and analyze the physiological stability index PSI with the first threshold Q1 to obtain the first evaluation result, including: When the physiological stability index PSI ≥ the first threshold Q1, it indicates that the user's physiological state during sleep is stable and there is no risk of triggering snoring due to abnormal oxygen deficiency and heart rate fluctuations, and continuous monitoring is carried out; When 60% of the first threshold Q1 ≤ the physiological stability index PSI < the first threshold Q1, it indicates that the user's physiological state during sleep is at the first level of abnormality and there is a risk of triggering snoring due to abnormal oxygen deficiency and heart rate fluctuations. Trigger the first warning instruction and generate the first strategy: use a tongue position muscle movement device to apply low-frequency electrical stimulation to the back of the tongue at 30Hz - 50Hz, 2 seconds each time, so as to reduce the heart rate value by 10% and increase the heart rate variability value by 15%; When the physiological stability index PSI < 60% of the first threshold Q1, it indicates that the user's physiological state during sleep is at the second level of abnormality and there is a risk of triggering snoring due to abnormal oxygen deficiency and heart rate fluctuations. Trigger the second warning instruction and generate the second strategy: monitor the user's breathing condition.

6. The intelligent oral snore stopper system based on multi-parameter monitoring according to claim 5, characterized in that, The airway perturbation monitoring module includes a second calculation unit and a second analysis unit; The second calculation unit is used to monitor the user's breathing condition in real time when receiving the second warning instruction. After dimensionless processing in combination with the data in the structure dataset, the breathing interference coefficient HXY is calculated and obtained. The formula is as follows: Wherein, DTC represents the value of the posterior displacement offset of the sound source, Zio represents the impedance ventilation patency, Jti represents the muscle tension value, Ymc represents the throat compression ratio, RR represents the real-time respiratory rate, and RR opt represents the individual optimal respiratory rate, DRR represents the number of deep breaths per minute, and DRR max represents the maximum expected number of deep breaths per minute.

7. The intelligent oral snore stopper system based on multi-parameter monitoring according to claim 6, characterized in that, The second analysis unit is used to preset a second threshold Q2 in advance and compare and analyze the breathing interference coefficient HXY with the second threshold Q2 to obtain the second evaluation result, including: When the breathing interference coefficient HXY < the second threshold Q2, it indicates that the user's breathing state during sleep is stable and there is no risk of triggering snoring due to structural throat compression or abnormal ventilation function, and continuous monitoring is carried out; When 130% of the second threshold Q2 > the breathing interference coefficient HXY ≥ the second threshold Q2, it indicates that the user's breathing state during sleep is at the first level of abnormality and there is a risk of triggering snoring due to structural throat compression or abnormal ventilation function. Trigger the third warning instruction and generate the third strategy: start a micro motor to gently push the tongue body forward by 2mm; apply a slight vibration stimulation through a flexible diaphragm to activate the sublingual muscles; emit a slow guiding breathing rhythm audio through bone conduction; When the respiratory interference coefficient HXY ≥ 130% of the second threshold Q2, it indicates that the user's respiratory state is abnormally level two during sleep, and there is a risk of triggering snoring due to structural throat compression or abnormal ventilation function, triggering the fourth warning instruction and generating the fourth strategy: monitoring the user's sleeping posture status.

8. The intelligent oral snore stopper system based on multi-parameter monitoring according to claim 7, characterized in that, The sleeping posture monitoring module includes a third calculation unit and a third analysis unit; When receiving the fourth warning instruction, the third calculation unit is used to monitor the user's sleeping posture status in real time. After dimensionless processing in combination with the data in the structural data set, the sleeping posture airway compression coefficient PSRF is calculated and obtained. The formula is as follows: In the formula, TG represents the tongue pressure of the user, α represents the head elevation angle, γ represents the mandibular elevation angle, DTC represents the posterior displacement offset value of the sound source, Zio represents the impedance ventilation patency, and β avg represents the average head offset angle, and β ref represents the normal posture reference angle.

9. The intelligent oral snore stopper system based on multi-parameter monitoring according to claim 8, characterized in that, The third analysis unit is used to preset a third threshold Q3 in advance, and compare and analyze the sleeping posture airway compression coefficient PSRF with the third threshold Q3 to obtain the third evaluation result, including: When the sleeping posture airway compression coefficient PSRF ≤ the third threshold Q3, it indicates that the user's sleeping posture structure is within a reasonable range and there is no risk of airway compression, and continuous monitoring is carried out; When the sleeping posture airway compression coefficient PSRF > the third threshold Q3, it indicates that the user's sleeping posture structure is outside the reasonable range and there is a risk of airway compression, triggering the fifth warning instruction and generating the fifth strategy: using a micro pneumatic adjustment structure to adjust the mandibular posture by 2 - 5 degrees, and performing non-awakening intervention with an ultra-low sound pressure < 30dB audio and a local bone conduction vibration frequency of 150 - 250Hz. The initial vibration duration is 0.8 seconds. If the user does not adjust the sleeping posture, it is extended to 2 seconds, and the maximum continuous stimulation does not exceed 5 times.

10. The intelligent oral snore stopper method based on multi-parameter monitoring, including the intelligent oral snore stopper system based on multi-parameter monitoring according to any one of claims 1-9, is characterized in that, It includes the following steps: Step 1: Collect the user's tongue pressure, oral temperature, oral muscle tone, heart rate and heart rate variability, blood oxygen, respiratory rate and deep breathing times to construct a physiological data set; collect the head and mandibular posture parameters, sound source offset, air flow smoothness and throat compression ratio to construct a structural data set; Step 2: Construct an initial model through a convolutional neural network, and use the physiological data set and the structural data set as inputs for training and testing; use the trained model for snoring state evaluation and intelligent intervention decision-making, extract the intermediate layer feature vectors to identify the deep features, and continuously optimize the model to achieve dynamic prediction of the snoring risk of different individuals; Step 3: Monitor the user's physiological condition in real time, calculate and obtain the physiological stability index PSI in combination with the data in the physiological data set, and compare and analyze it with the first threshold Q1 to judge whether the user's physiological state is abnormal during sleep. If it is abnormal, a strategy is given; Step 4: Monitor the user's respiratory condition in real time, calculate and obtain the respiratory interference coefficient HXY in combination with the data in the structural data set, and compare and analyze it with the second threshold Q2 to judge whether the user's respiratory state is abnormal during sleep. If it is abnormal, a strategy is given; Step 5: Monitor the user's sleeping posture status in real time, calculate and obtain the sleeping posture airway compression coefficient PSRF in combination with the data in the structural data set, and compare and analyze it with the third threshold Q3 to judge whether the user's sleeping posture structure is within a reasonable range. If it is outside the reasonable range, a strategy is given.

Images